An Alternative Approach to Troubleshooting Variable Peak Height/Area in HPLC
When a problem with an HPLC separation occurs it is often best to take a step back and think about what may have caused the problem, before taking any drastic action such as dismantling the system and then finding that you don’t know how to put it back together, or that once you do, the problem still persists. An Ishikawa diagram (also called a Fishbone diagram) is a useful tool that can be used to aid in troubleshooting. They are used to show the causes of a specific problem. The branches of the diagram that are more highly populated may have a greater influence on the final problem.
The problem of variable (irreproducible) peak height/area is a common and complex one, with many potential causes. We have constructed an Ishikawa diagram to look at all the possible causes of variable peak height/area and to help prioritise the approach to fixing the problem. Use the interactive diagram below to explore this more, or download the PDF as a useful reference tool for your lab.
Unfiltered mobile phase
Here we need to consider the actual pore/frit size which is recommended for the various stationary phase particle size/system combinations.
<3 µm particle size packing material – use a 0.2 µm porosity filter
>3 µm particle size packing material – use a 0.45 µm porosity filter
Contamination from wash bottle cap
When a needle wash bottle or vial is used, this should remain uncapped to prevent cross contamination from the cap itself
Analyte or sample matrix absorption onto metal or plastic surfaces
Autosampler contamination usually manifests itself as poor quantitative reproducibility, carryover (peaks in a blank injection which is made following a sample) or ghost peaks.
It is important to ensure that the sampling rate is correct, especially when working with Fast HPLC techniques to ensure that enough points are recorded across each peak for proper ‘modelling’ of the peak area. Generally 20-25 points across a peak are required for accurate quantification. Having an incorrect sampling rate can affect the peak shape, and if the rate is too low the peaks may look broad and one might interpret this effect as low column (system) efficiency. A very high sampling rate can lead to the baseline appearing very noisy
Most flow cells can be cleaned with water, warm water, methanol or other solvent (0.1N nitric acid solution is sometines used in extreme circumstances) but be sure to follow manufacturer’s instructions.
This issue mostly affects operation in gradient analysis mode. Typically called the ‘gradient dwell volume’ the pump dead volume affects the time taken between the mixing of the solvents within the pump and the delivery of those solvents to the HPLC column (along with the internal volume of system tubing and the autosampler volume).
Typically the problem can be solved by reverse flushing the column (disconnect from the detector first) in a solvent which will either flush away trapped particulate materials or dissolved absorbed contaminants, in which case one must choose a flushing solution in which the suspected contaminants are highly soluble.